The Found Corporation

Ubuntu One: Where Do I Look? What Do I Do?, a photo by pjsherman on Flickr.

This is an object lesson in how not to lay out a screen.

Background: While configuring a Linux PC, I found myself setting up Ubuntu One, Canonical’s cloud sync service. After I entered the confirmation code I received via email, I was taken to this screen.

My first reaction: Blargh! Where do I look first? What do I do?

Let’s break it down:

1. None of the calls-to-action appear to be primary. Nothing screams “Start here! Click me first!” The most visually prominent calls just sort of step on each other.
2. The layout is an eye-hurt. No really. I’m not exaggerating, it /actually/ hurts my eyes. I can’t recall the last time a screen layout actually made me go cross-eyed. Little thought seems to have been given to aligning the controls, text and links. I don’t mean to be too snarky…but it looks like the UI fairy barfed up links and controls onto the screen.
3. When I was a wee UX lad, mama always told me “If you can’t choose one font size, might as well use them all.” Unfortunately, mama was wrong. On this screen it’s disorienting. It adds to the difficulty users will experience trying to find a “start here” call-to-action.
4. Screenshots are helpful…in theory. But they have to illustrate something helpful to the user. This one falls short. The detail is too small to be of use. It just causes eyestrain. Or maybe I’m just getting old, and you youngsters have no problem with it. Well good for you. And get off my lawn.

I know I’ve ranked on Linux in the past, so I readily admit that Canonical has made great strides creating a better user experience for desktop Linux. Ubuntu 11.10 is by far the most usable Ubuntu release ever, and Ubuntu has always stood above other distros in ease of install, configuration and initial/ongoing use. But like any app or OS that comes from a legacy of “built by geeks for geeks”, there will always be areas of the user experience that don’t get the UX love and attention they deserve. This screen – even though it was probably created for more recent distributions – reminds us of Linux’s geeky origins.

One last thing, and this is a completely preference-driven personal peccadillo of mine…bevel-lowered grouping boxes must die. Hate ‘em. They’re /so/ Windows 3.11…

This is an impromptu video of me interacting with the Google+ “add people to circles” feature.

Like most things we interact with in the world, there’s some good and some not-so-good to the experience.

First, the good:
1. The people “cards” are just the right size. The picture is just right also – big enough so I can recognize the person, small enough to fit their name.
2. It was also a smart idea to just show the target’s name (or email if the name is unavailable). It makes for easier and quicker scanning and target acquisition.
3. The whole card is a click target.
4. You can select multiple cards with clicks; no CMD+ or shift+click necessary. However, using a key combination (which some people will carry over from the desktop software interaction paradigm) works just fine as well.

The not-so-good:
1. Drag and drop appears to be the only way to get people into circles. Really? Given Google’s demonstrated commitment to accessibility, I *can’t* believe this is the only way to manage circle assignment. I must be missing something. That just can’t be.
2. A few times I highlighted multiple cards and then inadvertently deselected them. I have no idea why or how.
3. Believe it or not, on first view I thought Google+ had only found 28 (7 x 4) people I knew. It took more time than it should’ve to realize that the card area was scrollable. I might not be able to pin that on Google however. I used Safari on Mac OS X 10.7, which together are a walking advertisement for iOS disappearing scrollbars.

So what does this little spiel demonstrate? I guess it shows that even cutting-edge design with massive resources behind it can still occasionally be a struggle for users. Oh, and don’t forget the accessibility.

I’ve Found It: The Worth Thermostat Ever, a photo by pjsherman on Flickr.

After many years of working in user experience, I’ve found the semi-mythical holy grail of poor design: in my grandmother’s condo I encountered the dreaded “which way is which?” thermostat.

Just looking at this picture, you might think that pressing the left-facing button would lower the temperature, and pressing the right-facing button would increase it. And you’d be wrong. At least, I think you’d be wrong.

If only they were reversed, I could at least deal with that….but sadly no, what we have here is *both* a 90-degree control rotation (or negative 90; I’m not sure), and some kind of mode problem where I couldn’t actually just change the temperature on the fly without entering a “temperature change” state.

And no, I never discovered how to enter the “change temp” mode…I just sweltered in the Florida heat for the rest of the visit.

Here’s a great example of how feature requirements and design have combined to cause a lost sale. Some background: This is the donation page for the Presidential candidate I support. I’m not getting into who it is; those who know me personally can probably guess, and those readers who don’t can visit candidates’ donation pages until they find this example.

I’m your typical semi-involved voter: I care a bit about local policitics, I try to vote in every election, but don’t always make the smaller or off-year ones. But I do care about state and national politics, and since I live in *the* quintessential swing state, I feel like my vote actually matters.

This morning I received a call-to-action email asking me to donate, and followed the link, landing at the page you see (in part). I noticed the “Make this a monthly recurring donation” checkbox, and immediately thought “Awesome! That’s what I want to do!” I figured it would be easier for me to just automagically kick in $25 per month until the general election was over. So I ticked the box, expecting to get one of those “For how many months?” dialog or AJAX partial page update. Nope. Nothing. Which sucks for the candidate and party, because I – and I suspect many, many other small donors – only want to make a recurring donation until the general election is over. And if I can’t do this – or easily figure out how to do this – there is no way I’m setting up an autopayment.

This is a classic failure of the product managers and designers to understand the customer. They may *want* me to set up an autopay in perpetuity. But I guarantee that most small, semi-involved donors most certainly do not want to do this.

Lesson: Learn what your customers want from your product and what their motivations are when they use it. Wishful thinking about user behavior is not a good way to meet their wants and needs.

I’m on my way back from my company’s all-hands meeting.

After all the excitement and motivation the week inspired, I felt moved to write a UX mini-manifesto. A mini-festo, if you will.

Excuse any grammar or spelling issues; I’m composing in Evernote on my phone. I would love to hear my readers’ and followers’ comments on this post.

You’re an experience design practitioner. In your organizations, you should be responsible for:

• Creating an inviting and well-designed initial user experience.
• Designing and validating:

Along the way, you also:

• Collaborate on the definition and optimization of product development lifecycle processes with our functional neighbors – i.e. PM, Dev, QA, Marketing, and Social/Community Management.
• Measure, track and improve the user experience.
• Discover opportunities to delight customers in ways that are not easily discoverable by market-level research methods.
• Occasionally uncover strategic jobs that customers need doing, and design opportunities for more sustaining vs. incremental product innovations.
• Provide the business with both strategic and tactical customer insights and understanding.

eBay: Hah Hah, Made You Think!, a photo by pjsherman on Flickr.

Standard disclaimer: I am a user and fan of eBay. When it comes to user experience, they do lots of things right.

Here’s one thing they did wrong: They provided half-hearted, linkless “help” in the form of “to do x, go to [place A] or [place B]“, without including links to those locations. This is a no-brainer and should’ve been coded ages ago.

As a result, I had to hunt around for a small but still-annoying period of time before I found where I needed to go.

Somebody add that to the eBay UX fix list.

Message Can’t Be Blank? Why The Hell Not?, a photo by pjsherman on Flickr.

This screenshot is about a year old, so my apologies to the site if they’ve fixed this UX issue.

The point remains, however, that when someone wishes to unsubscribe from your email list, it is incredibly imperious to *require* them to explain why.

Sure, as a business you might *like* to know, but you have no right to demand an answer.

That’s treating your visitor rudely.

A Dubious Claim, a photo by pjsherman on Flickr.

Paypal goes a little too far with this login transition screen.

“The world’s most loved way to pay and get paid”?

I don’t think so.

Using marketing-speak in the UI is one thing. It’s come to be expected and (mostly) accepted. Just make sure it doesn’t trigger your users’ BS detectors.

Whatever You Do…Don’t Search For “ASCII Mr. Burns” Or Google Will Flip Out, a photo by pjsherman on Flickr.

Really Google? I’m on my home network. No one else is using my connection, according to my outbound connection monitoring software.

Nothing but little ol’ me on Firefox.

One-Try Lockout? That’s BS., a photo by pjsherman on Flickr.

This is a pure customer experience issue: I was just locked out of my AT&T wireless account after a SINGLE INCORRECT PASSWORD ENTRY.

I suppose it’s theoretically feasible that someone was trying to crack my account at the very moment I was trying to legitimately log in, which would explain why I was locked out after a single try. Odds are, probably not.

Thanks the inconvenience, AT&T.

And companies wonder why we call instead of self-service on the web. It’s because the web is basically a brittle support mechanism.

Given that the average support call costs anywhere between 5 and 50 dollars (sorry, no recent reference; going from memory), I’m about to cost AT&T ~ 15 or 20 bucks because someone decided to implement a single-attempt lockout policy.

Whatever. Time to call a CSR and burn some of AT&T’s cash.